ClawHub
Back to skill

Security audit

MLOps Observability

Security checks across malware telemetry and agentic risk

Overview

This skill is a coherent MLOps observability helper, with a normal caution that MLflow logging can expose training data details if pointed at a remote server.

Before using this skill, review the MLflow tracking destination. Keep it local or trusted when training data, model artifacts, or git metadata are sensitive, and avoid logging raw or regulated datasets unless your MLflow storage and retention policies are appropriate.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The code reads a dataset from a configured path, converts it to an MLflow dataset object, and logs it to the tracking backend without any consent check, minimization, or warning about where the data will be stored. In an MLOps/observability skill, this is especially relevant because MLflow may be configured to use a remote server, making accidental transmission or long-term storage of sensitive training data more likely.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal